Endovascular Therapy after Stroke: A Penumbra of Doubt

Published - Written by Rena Xu

In the acute window after an ischemic stroke, you have two options. You can give tissue plasminogen activator (t-PA) to dissolve the clot, or you can use endovascular therapy to remove the clot. The probability of recanalization is 46% with t-PA, and greater than 80% with endovascular treatment. Which option do you choose?

It might be tempting to formulate a preference based on these figures alone; the problem, however, is that a higher recanalization rate does not necessarily translate to superior clinical outcomes. In fact, several studies published recently in the Journal suggest there may not be a difference in outcomes between the two options.

Kidwell et al. report the results of MR RESCUE, a trial that randomized 118 acute stroke patients to receive intravenous t-PA (standard care) or undergo mechanical embolectomy within 8 hours of symptom onset. All patients underwent pretreatment brain imaging to determine whether they had a penumbral (favorable) or non-penumbral (unfavorable) infarct pattern. It was hypothesized that patients with a penumbral pattern, indicating an area of salvageable brain tissue, stood to benefit selectively from revascularization.

This is not what the results demonstrated. Overall, patients with a penumbral pattern did have slightly better outcomes as compared to those with a non-penumbral pattern(mean modified Rankin score was 3.6 vs. 4.2 on a scale of 0 to 6, where a lower score indicates less disability; P=0.047), but this was independent of treatment assignment. Endovascular therapy did not favor those with penumbral as compared to non-penumbral patterns (3.9 vs. 3.4; P=0.23). More broadly, endovascular therapy did not demonstrate clinical superiority to medical therapy with t-PA (3.9 vs. 3.9; P=0.99).

Two other studies offer evidence to support this last finding. Ciccone et al. report the results of a randomized controlled trial that compared endovascular therapy to intravenous t-PA. The trial enrolled more than 360 patients and looked at outcomes at 3 months. No significant difference was observed in disability-free survival between the endovascular therapy and intravenous t-PA groups (30.4% vs. 34.8%; absolute difference of -4.4 percentage points, 95% CI: -14.1 to 5.2; P=0.37).

Finally, a study by Broderick et al. compared the effectiveness of treatment with a combination of endovascular therapy and intravenous t-PA versus t-PA alone. The trial, which enrolled more than 650 participants, was stopped early due to futility; no difference was observed between the study arms in terms of functional independence (modified Rankin scale score of 2 or less), mortality at 90 days, or rate of symptomatic intracerebral hemorrhage after administration of t-PA.

NEJM editor Mary Beth Hamel, M.D., states: “These findings highlight the importance of randomized controlled trials, with appropriate control groups and relevant clinical outcomes, in assessing the clinical value of medical devices.”

The findings also challenge us to reconsider any preconceived notions of therapeutic efficacy. Kidwell et al. point to the “relative lack of equipoise in the stroke community regarding the putative benefits on clinical outcomes of embolectomy versus standard medical care” as a barrier to designing and executing randomized clinical trials. The greatest contribution of MR RESCUE may be to lower this barrier and pave the way for future comparative effectiveness studies. As a starting point, it would be valuable to replicate the study using newer generation embolectomy devices (MR RESCUE used first generation devices). Further work is also needed to tease out the decision support value of pre-treatment imaging.

The options for acute stroke treatment are few, but until more is known, the choice between them will remain anything but straightforward.

How do imaging findings factor into your treatment decision for acute stroke patients?  How will the results of these studies affect your use of endovascular therapy versus intravenous t-PA?